When I first started using SO about 2 years ago, there were lots of questions, and lots of answers. Great answers even. It was wonderful because you were almost guaranteed to get some kind of answer.

Now it seems like getting even 1 answer is increasingly difficult. I find that many questions are by people who have question to answer ratio's of 10:1 or more (mine is 1:15).

Would it be possible to send a gentle reminder to people when their ratio goes above say 5? I don't think it would be productive to force them to answer more questions (who wants bad answers), but some kind of encouragement to share their knowledge.

Am I alone it seeing this disappointing trend?

  • 5
    Just in case you aren't aware of it, but there are measures in place in the case of people who ask too many poor quality questions. You can read up on it over here.
    – Grace Note StaffMod
    Sep 27, 2010 at 20:54
  • 1
    1:15? You ask way too many questions, friend.
    – Welbog
    Sep 27, 2010 at 21:35
  • 1
    @toronto 1 question per 15 answers... or am I missing a joke here? Sep 27, 2010 at 21:48
  • 5
    @Toronto - It doesn't matter how many questions you ask, as long as they're good. I'd rather have a user with 100% questions that are good. Than a user with 5% questions, but those questions are bad...... bad as in unclear and not useful. Sep 27, 2010 at 22:19
  • If you look through the unanswered questions, they are basically trash. For example [ here are the unanswered and open jQuery questions ](stackoverflow.com/search?q=answers:0+[jquery]+closed:0). It's very difficult to find a good question in there. The good questions get a lot of attention. There are more and more bad questions as the number of users increases. ----------------- Aarobot does have a good point that it should be much easier to find the few good questions hidden in all those bad questions. Sep 27, 2010 at 22:24
  • I've noticed this trend too, especially since I MOSTLY ask rather than answer. On a personal note, though: if you want to get a question answered, time-of-day is the most important determinant of the quality and quantity of answers you'll get. See my question here meta.stackexchange.com/questions/7105 NOTE: this is my personal opinion and may not necessarily agree with the facts ;) Sep 28, 2010 at 5:55
  • Byron, one question per fifteen answers is one question too many.
    – Welbog
    Sep 28, 2010 at 13:47

6 Answers 6


I don't think the problem is that people are refusing to share their knowledge. The problem is that the oldest question on the front page is 12 minutes old. These days, if your question doesn't get immediate attention, then it has almost no chance of being answered.

I've been saying this for a long time, but I don't think the team believes me because of Stack Overflow's very well-deserved success so far. They think it's an absolutely sensational idea to put the entire programming community under one roof. I think that it was an absolutely sensational idea, but it doesn't scale very well to 300,000 users or whatever the current count is - at least not with today's design.

Sure, you can use interesting/ignored tags to filter questions, but that doesn't really actually filter them, it just highlights a few questions and hides the other unwanted ones from your page. If you choose to ignore [.net] and [java] and maybe a few other tags, you'll just end up with an empty front page most of the time, and the team has declined requests to improve this (i.e. to actually show the same # of questions, but filtered server-side).

So what we end up with is a lot of unanswered questions, because everything moves way too fast for people to find all the questions that they're interested in and capable of answering.

It is my opinion that there is no possible way for the community itself to solve this problem; the solution needs to come in the form of site improvements, and those are (in descending order of preference):

  • Vastly improved ability to customize the front page, so as to allow users to see only the types of questions they are interested in and still see an appropriate number of them. At a minimum, this would require people to be able to show only their interesting tags (hide all other questions) with server-side filtering. Better sorting heuristics would also help, i.e. to rank based on view count, votes, hide questions with accepted answers, etc.

  • Vastly improved slicing-and-dicing of the unanswered questions section. This would include all of the same aspects as the front-page customization - proper filtering and more advanced sorting, as well as some additional features tailored specifically toward distinguishing the bad questions from the hard questions: Filtering/sorting/highlighting based on things like: The author's accept rate and reputation, their average/min/max question score, close votes, comments (i.e. to show that the author responds), etc. Questions like this one are nothing but noise.

  • Better incentives for answering unanswered questions. The necromancer badge is fun but hardly worth the effort of slogging through hundreds of old questions - especially knowing that they'll get low views and hence low votes. These questions are often very difficult to answer with little to nothing in the way of rewards. And the new Tenacious / Unsung Hero badges are all but unattainable if you're like me and already have hundreds of upvoted answers.

  • Better incentives for downvoting and janitorial work. Many of the unanswered questions should be closed and deleted or at least massively downvoted. Getting them removed from the unanswered question page would be great. Getting them punted off the front page would be even better. As it is, a lot of the really crap questions come from low-rep users or random Google hits so it feels silly to downvote them - and it's a waste of a close vote because not enough other people will see them. (If more 10kers used the tools to identify close candidates, that would help - again, incentives.)

  • And there was something else that I can't remember anymore.

One of the main reasons I cut down heavily on Stack Overflow time was that it was just getting too hard to keep up. There are so few questions of interest to me on the front page at any given time that most of the time, when I did deign to answer one, I had to abandon it shortly thereafter because somebody else beat me to the punch. I'm not complaining about FGITW - I can answer pretty damn fast - the problem is that every single person is focused on the same 5 or 6 questions.

It's not that there aren't any interesting questions. I just can't see them.

If people had better control over their view, different people would be looking at a different subset of questions. While everybody else is hammering away at the insipid interview question I could be crafting a detailed answer to the obscure WCF REST question that was posted 45 minutes ago and everybody else ignored.

And of course I realize that these are all "power user" features that 95% of the members would never use. To me, that's okay, because the other 5% of the members are probably contributing about 50% of the answers. (Okay, I made up that statistic, but I'll bet it's not far off.)

There's no low-tech human solution to this problem. The incentives and UI design are all biased toward getting new and easy questions answered, and maybe that is a good thing, but I think we've reached the scale where we need to give experienced users the opportunity to make their mark by answering meatier questions instead of everyone attacking the beginner question that was posted 2 minutes ago.

And that opportunity needs to part of the site design - it's not going to magically appear if we start telling people to answer more questions.

  • 13
    +1 very good analysis. I've been cutting way back on my SO usage for the same reasons. It is a shame since it is going to do nothing but hurt the community. As more the experienced veteran programmers stop visiting, the site becomes less and less useful. Without some changes (like you mentioned) It could easily become a shell of it's former self. I hope SO doesn't become a victim of it's own success. Sep 27, 2010 at 22:08
  • Also what are some of the tools we can use to identify close candidates? Maybe I am missing something. Sep 27, 2010 at 22:11
  • @Byron: If you are 10k, click "tools" on the top and then go to the "close" tab, you'll see questions with existing close votes. They've actually improved that screen, it used to list only 5 questions at a time.
    – Aarobot
    Sep 27, 2010 at 23:01
  • +1 Nicely explained and some good ideas in there. I agree that the system needs to grow and evolve, as well as the community. And I do know that SO is under constant dev (codinghorror.com/blog/2010/09/go-that-way-really-fast.html) but it could be time for some major tweaks.
    – Dhaust
    Sep 27, 2010 at 23:13
  • 1
    I never use the front page. It's pretty boring, only 25% of the questions there are on topic for me. I just run my own query, selecting by tag. That's not common? Sep 27, 2010 at 23:30
  • @Hans: Sure, you can do that, but it's rather inefficient and if you're active in a big area like .NET then either you're still going to get a lot of noise by querying on the main tag, or you'll need a dozen different queries for all of the more obscure sub tags. Much of what I'm asking for is simply a better way to query; I have a pretty good idea of how to find the questions I'm interested in, but the system just doesn't let me do it efficiently.
    – Aarobot
    Sep 28, 2010 at 0:48
  • Why would an experienced user want to use the front-page view? Use the "newest" and turn up the answer count. Sep 28, 2010 at 0:59
  • I rarely use the front page. I check the first 10, then goto [java] and look at 2-3 pages of questions I actually have a chance of answering. I can understand though that it might be hard if your following a massive tag like [c#]
    – TheLQ
    Sep 28, 2010 at 2:13
  • @dmckee: Splitting hairs here. The difference is marginal. The point is that new questions roll in far too fast for all good questions to get answered by people only looking at the newest questions.
    – Aarobot
    Sep 28, 2010 at 2:39
  • Agreed that better tools / incentives are required for janitorial work. I've almost given up on marking duplicates because very rarely do enough people with the power to vote see them in time to close them. Perhaps a 10k tool option to see open questions with at least one vote as a duplicate would help, because even if it's not your field of expertise it's usually easy to spot the similarities between questions once they've been pointed out. Sep 28, 2010 at 18:35

When SO was still shiny and new, the more experienced and serious programmers were asking every good question they could think of, drowning out all the "plz send teh codez" questions. Now we're mostly left with the latter.

I think SO has reached a certain plateau for mainstream programming questions. Basically, everything worth asking has been asked. As technology changes, we'll see spikes of new questions (imagine if MVC or LINQ were brand new today).

Further, since all the obvious questions are well covered, all we have left is obscure questions. This may be why you're having trouble getting answers. If no one has asked before, maybe your problem is unique to your situation. Don't let this stop you, though. It's always worth a try. (If you have any specific examples, we'll all be happy to take a look)

As Grace Note mentioned, there are limits in place that will throttle or even completely cut off users who ask too many [bad] questions. However, there certainly are plenty of these users to go around.

  • 1
    B, looking at the front page of SO now, how many BAD BAD questions do you see? I'm counting 0 out of 50 at the moment.
    – waffles
    Sep 28, 2010 at 2:05
  • Agree that tons more questions are obscure these days ... but I don't have numbers to back me up yet
    – waffles
    Sep 28, 2010 at 2:07

Here are some rough stats:

  • Jan 2009: approx 5 answers per question
  • July 2009: approx 3 answers per question
  • Jan 2010: approx 2.5 answers per question
  • Now: approx 2 answers per question

So yes there is defiantly a trend that as the site got busier we started getting less answers on average for questions.

There are some other interesting stats ... prior to September 2009, people were barely accepting answers to questions (1/5) we are holding steady at the moment with an approximate 50% accept rate on questions.

As to how to solve this problem:

Would it be possible to send a gentle reminder to people when their ratio goes above say 5?

I think this is a really bad idea. As others have said, there are already limits in place that stop people from asking a stream of poor questions. In fact when I look at the front page these days I notice very few (if any) extremely poor questions, this was not the case 2-3 months ago.

A Q&A site without great questions is useless, if somebody is asking tons of great questions we do not want to tell them: "hold on buddy, time to answer some questions", instead we would like to tell them: "wow, you are asking awesome questions ... keep it up"

If you look at tag pages, like ruby sql or css you start seeing questions spanning a whole day on a single page, unlike the front page, that at peak times can only cover 20 minutes.

I agree that we want to make it easier for people to answer questions, I am not sure exactly what to do to make it so. Its a large combination of adjustments.

Our ultra-high rep users are super proficient at finding and answering really interesting questions and they manage to do so regularly, daily.

When you filter down to a 5 hour time span in the c# tag you start noticing that the unanswered questions are actually pretty tough, they involve some sort of nitch technology and there are very few people who can answer them. All the easy questions seem to get answered.

From personal experience, I tend to answer my own questions if they do not get a satisfactory answer. I guess there are tons of cases where the OP has solved some really hairy problem but never bothered updating the site.

So what should we do:

  1. Make it easier for people to find questions they can answer.
  2. Encourage closing the loop, find better ways to remove old unanswered questions from the unanswered question list.
  3. I personally think that a system that encourages collaboration (provide an incentive for improving questions and answers) will do wonders at clearing up a lot of the old unanswered list, quite often there is a great question hiding in a poorly worded one.
  • 1
    Is it possible that the reduction in answers/question actually means that people are getting higher quality answers?
    – Jon Seigel
    Sep 28, 2010 at 1:41
  • @Jon accept rates are pretty stable, so I think the quality is holding. I would expect a reduction in low quality questions would actually reduce answers/questions cause low quality questions and polls tend to get tons of answers. I will need to re run stats and exclude wiki / polls and closed stuff.
    – waffles
    Sep 28, 2010 at 1:44

While I agree that it's becoming a problem do you really want the guy that asks 5 questions a day to answer your question? =)


I only look at the front page for fun, bikeshed questions. The rest of the time, I use the tag for my programming language (ruby). Apart from one or two problem users, I haven't noticed things get better or worse.


Other than the ways of dealing with problem users, there's little that one can do to affect how people choose to use the site. If you're having a problem getting answers to your questions, you can help yourself. If you can make even a weak case for putting one of the popular tags (e.g., C#) on your question, do it. That will get you enough initial views to get some good answers, even if the question is re-tagged later.

  • 5
    Do not encourage people to put tags on their questions to get more attention. Such behavior is rude. Not a little rude. Not sorta-rude-but-it's-OK-if-that's-what-it-takes. Just plain, old-fashioned, "I'm more important that you" rude because it constitutes a claim that the asker's desire for an audience supersedes the potential answerer's desire to control what they spend their time on. I consistently downvote any question that seems to have engages in such gaming. Sep 28, 2010 at 0:57
  • 1
    Please no, I'm tired of [c#] questions that get shotgun tagged with [java], [php], or other nonsense.
    – TheLQ
    Sep 28, 2010 at 2:18
  • @dmckee, WOW. You find that rude? I'm not talking about lying in your tags - I mean finding a way for your question to also be about C# so you can use the tag. If a question doesn't have one of my Interesting tags, I won't see it. There are far too many questions to read them all. If you can't get your question some attention while it's close to the front page, you odds of getting good answers go way down. And if that represents a problem with the site, that's fine. That's not gaming the system. It's finding a way to make your question interesting to your audience. Sep 28, 2010 at 4:30
  • @dmckee, @TheLQ, if we need to reword or delete my answer to make it less bad for the community, let's do so. I'm in good faith not trying to encourage behavior perceived by everyone to be rude. Sep 28, 2010 at 4:31
  • So your advice is "Ask something close to what you mean so you can put a popular tag on it." Then what? Use the comments to advise your new larger audience about what you really meant? Ignore ansewers to the question you asked in favor of answers to the question you meant? Again, you are prioritizing your desire for a large audience over everyone else's attempts to use their SO time efficiently. Ask the question you have, tag it correctly. Full stop. Sep 28, 2010 at 14:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .